I've been re-reading One Useful Thing while writing a review of Co-Intelligence, Ethan Mollick's new book. Like you, and Alison Gopnick, Henry Farrell, and Brad Delong, Mollick points to how generative AI is at its core a cultural technology. "Magic for English Majors" is a good essay on this.
For all the supposed mystery of how LLMs are coming up with such good facsimiles of human thought or images, they are just just black boxes running statistical programs sitting on top of the internet. "Reprocessors of what has been said and made" is a good description. If you want value from using these reprocessors, as you say, you'd better know something about the history and social contexts of the steaming lumps of culture the machines regurgitate.
If only we could gather together people deeply knowledgeable about culture and history and have them offer educational experiences to young people interested in learning about culture and history, maybe we could...Nah! Let's spend trillions on building general intelligence, something Stephen Jay Gould long ago explained does not exist.
I've been re-reading One Useful Thing while writing a review of Co-Intelligence, Ethan Mollick's new book. Like you, and Alison Gopnick, Henry Farrell, and Brad Delong, Mollick points to how generative AI is at its core a cultural technology. "Magic for English Majors" is a good essay on this.
For all the supposed mystery of how LLMs are coming up with such good facsimiles of human thought or images, they are just just black boxes running statistical programs sitting on top of the internet. "Reprocessors of what has been said and made" is a good description. If you want value from using these reprocessors, as you say, you'd better know something about the history and social contexts of the steaming lumps of culture the machines regurgitate.
If only we could gather together people deeply knowledgeable about culture and history and have them offer educational experiences to young people interested in learning about culture and history, maybe we could...Nah! Let's spend trillions on building general intelligence, something Stephen Jay Gould long ago explained does not exist.